Visual Odometry

Research Project Goal

We a large corpus of panoramic video that is geo-referenced. The video has been collected by a car instrumented with high-recision GPS and inertial sensors. The goal of this project is to use visual sensing to compute the motion estimates between consecutive views and integrate it across the entire trajectory. The final results will be compared with the pose estimates obtained by GPS and intertial sensors.

Research Project Scope

Given a stream of images, compute the trajectory of the mobile platform.

Data

Video stream of the images acquired by a panoramic camera will be available.

Tasks

• Extract feature points
• Implement least squares solution to motion estimation between two views.
• Implement least squares solution to motion estimation to 3D structure recovery.
• Implement motion prediction step given the available structure.
• This will be followed by a windowed bundle adjustment over multiple frames. Integrate camera model of panoramic camera with calibration parameters provided by instructors.

Research Project Status

student names here

Point of Contact

Jana Kosecka

Midterm Report

not yet submitted

Final Report

not yet submitted